You are on page 1of 10

11th ICSGE Ain Shams University

17-19 May 2005 Faculty of Engineering


Cairo - Egypt Department of Structural Engineering

Eleventh International Colloquium on Structural and Geotechnical Engineering

DESIGN OF MONOPILE FOUNDATIONS FOR OFFSHORE


WIND ENERGY PLANTS

MARTIN ACHMUS1, KHALID ABDEL-RAHMAN2

ABSTRACT

The foundation of offshore wind energy plants plays an important role in the stability of these
structures. One foundation concept which can be used in this field is the monopile concept.
Here the foundation consists of one large-diameter pile (up to 8.0 m). Common design
practice, e. g. according to API regulations, does not cover horizontally loaded piles of such
dimensions. Thus, the soil-structure-interaction and the behaviour of these foundations has to
be covered using numerical modelling. This paper aims to investigate the behaviour of the
monopile under monotonous loading taking the interaction between the pile and the subsoil
into account. A three-dimensional numerical model using the finite element method was
established. In this model the non-linear material behaviour of the subsoil is described using
an elasto-plastic constitutive model. The interactions between the monopile and the
surrounding soil are modelled using contact elements. A parametric study with different pile
geometries, soil and loading conditions has been carried out. The results of the finite element
simulations are presented and evaluated. Finally, an overview on the state of knowledge
concerning the influence of cyclic loading is presented and open questions are discussed.

Keywords: Constitutive Model, Cyclic loading, Interface Friction, Monopile, Numerical


Modelling, Offshore Wind Energy Plants.

1 INTRODUCTION
The planned offshore wind parks in the German part of North and Baltic Sea will be
constructed in water depths varying from approximately 15 to 40 m. By means of suitable
foundation constructions, the large horizontal forces and bending moments resulting from
wind and wave loadings, must be economically and safely transferred to the sea soil. Mono-
pile foundations can be used as one of these foundation types. In principle the monopile is an
1
Prof. Dr.-Ing., Head of the Institute of Soil Mechanics, Foundation Engineering and Waterpower Engineering,
University of Hannover, Appelstr. 9A, D-30167 Hannover, Germany, Phone +49 511-762-4155;
achmus@igbe.uni-hannover.de
2
Dr.-Ing., Institute of Soil Mechanics, Foundation Engineering and Waterpower Engineering,
University of Hannover, Appelstr. 9A, D-30167 Hannover, Germany, Phone +49 511-762 2273;
khalid@igbe.uni-hannover.de
extension of the main tower into the soil under the sea bed (Fig. 1). This foundation method
was already implemented in North and Baltic Sea, but only for wind energy plants in water
depths of less than about 10 m. Its application is expected to be possible for water depths up
to about 25 to 30 m. The diameters of such monopiles vary between 6 and 8 m.

Since wind energy plants are relatively sensitive to deformations, in particular tilting, it is
really important to estimate these as exactly as possible. For the mentioned large-diameters
piles, till now there is no exact design procedure approved. In this paper, the results of
numerical investigations of the load-deformation behaviour of monopiles under static loads
are presented.
+ 190 m Soil profile

0m

-1 m S,u

+ 130 m

mS - fS

+ 30 m -7 m

+_ 0,0 m

mS,g

D = 7,5 m

Fig. 1: Example of a monopile foundation.

2 STATE OF THE ART FOR COMPUTATION OF HORIZONTALLY LOADED


MONOPILE FOUNDATIONS
The design procedure for OWEP foundations in Germany is given in the Germanische Lloyd
rules and regulations (GL 1999). In this regulation, concerning the behaviour of piles under
horizontal loading reference is made to the regulation code of the American Petroleum
Institute (API 2000). The Norwegian guidelines (DNV 2004) also refer to the API code. In
the API code the p-y method is recommended for the design of horizontally loaded piles.
In principle, the p-y method is a subgrade modulus method with non-linear and depth-
dependent load-deformation (p-y) characteristics of the soil springs.
API (2000) describes the construction of p-y-curves for soft and stiff clay as well as for sandy
soils. Due to API, p-y-curves for sandy soils can be derived as follows:

The maximum mobilized soil reaction force per unit length of the pile pu depends on
the regarded depth under sea bed z, the submerged unit weight of the soil , the pile
diameter D and on the angle of internal friction of the sand:
p us = (c1 z + c 2 D ) ' z (1.1)
pud = c3 D ' z (1.2)
The first mentioned equation applies to small depths (pus) and the second equation for
larger depths (pud), the smaller of both values is to be considered. The influence of the
internal friction angle is described by the factors c1, c2 and c3 (see Fig. 2 left).

The p-y-curve is described by the following equation:


kz
p = A p u tanh y (2)
A p u
with A = 3.0 0.8 z / D 0.9 for static loading
and A = 0.9 for cyclic loading.
Herein p is the soil resistance per unit length of the pile and y is the actual horizontal
deflection. The parameter k describes the initial modulus of subgrade reaction and is
dependent on the relative density ID and/or on the angle of internal friction (Fig. 2
right).

5 100 50

= 29

= 30
4 80 40

3 60 30
c2 Sand,
below G.W.L
2 40 20
c1

= 36

= 40
1 20 10
c3
0 0 0
20 25 30 35 40 0 20 40 60 80 100
Angle of internal friction in Relative Density ID in %

Fig. 2: Coefficients c1, c2, c3 and initial modulus k given in API (2000).

The equations (1) and (2) are based on investigations of Reese and Cox (Reese et al. 1974).
They tested a 21 m long steel tube pile having a diameter of 61 cm under different loading
and then evaluated their results. For cyclic tests, a maximum number of 100 load cycles was
realized. The correction factor A according to equation (2) was adjusted based on the
measurements done.

The application of this method worked satisfactorily in offshore practice over many years,
whereby the collected experiences only refer to piles with diameters up to 2 m. Therefore, the
API procedure should not be used directly for monopiles of large diameters (Achmus &
Abdel Rahman 2004, Lesny et al. 2002). After Lesny & Wiemann (2004) the subgrade
modulus for piles of large diameter is overestimated with the API method. They suggested a
diameter-dependent correction factor of the initial subgrade modulus k.
In the following, the results of numerical calculations of the load-deformation behaviour of
monopiles are presented and compared with the results of the API p-y method. The
calculations with the API method were carried out by means of the LPILE program (Lpile
2000).
3 NUMERICAL MODELLING OF MONOPILE BEHAVIOUR UNDER STATIC
MONOTONOUS LOADING

For the investigation of the load-deformation behaviour of monopiles of large diameters,


three-dimensional finite element calculations were accomplished. Piles with a diameter of D =
7.5 m having two different embedded lengths under the sea-bed of L = 20 m and L = 30 m
were investigated.
Different load application heights h of the load above sea-bed and thus combinations of
horizontal force H and bending moment M = H h were realized. Additionally, a vertical
load was applied to take the structures weight into account.

The computations were done with the program system ABAQUS (Abaqus 2004). In order to
carry out a lot of calculations for varying boundary and loading conditions, a large computer
system with parallel processor technology was used to minimize the time effort.
Due to the symmetric loading condition only a half-cylinder representing the sub-soil and the
monopile could be considered. The discretized model area had a diameter of 90 m, which is
twelve times the pile diameter. The bottom boundary of the model was taken 15 m below the
base of the monopile. With these model dimensions the calculated behaviour of the pile is not
influenced by the boundaries (Fig. 3 left).
For the soil as well as for the pile 8-node continuum elements were used. The frictional
behaviour in the boundary surface between pile and soil was modelled by contact elements,
whereby the wall friction angle was set to = 0.67 . The material behaviour of the
monopile was assumed linear elastic with the parameters E = 2.1105 MN/m2 (Youngs
modulus) and = 0.2 (Poissons ratio) for steel.
Of crucial importance for the quality of the numerical computation results of soil structure
interactions is the modelling of the material behaviour of the soil.
The elasto-plastic material law with Mohr-Coulomb failure criterion, provided in the
ABAQUS program, was used. This material law was extended in the elastic range by a stress-
dependency of the oedometric modulus of elasticity with the following equation:


ES = at m (3)
at

Fig. 3: Finite element mesh and horizontal bedding pressure (in the symmetry axis of the
monopile) for D = 7.5 m, L = 30 m in dense sand, H = 8 MN, h/L = 1.
Herein at = 100 kN/m2 is a reference stress and m is the current mean principle stress in the
regarded soil element. The parameter determines the soil stiffness at the reference stress
state and the parameter rules the stress dependency of the soil stiffness.
This material law has the advantage that it can be generally used for both cohesive and non-
cohesive soils. In the context of the computations presented here, the material parameters
used with reference to EAU (1996) are shown in Table 1.

Table 1: Material parameters used in the numerical computations.


Material Unit Stiffness Poissons Shear parameters
weight in 1 in 1 ratio in c in in
in kN/m3 in 1 kN/m2
Sand, medium dense 11 400 0.60 0.25 35 0.1 5
Sand, dense 11 600 0.55 0.25 37.5 0.1 7.5
Boulder clay, semi-firm 11 40 0.90 0.25 32.5 15 2.5
Mud sand 10 Es = 24 MN/m2 0.25 27.5 1 0

The finite element calculation is executed stepwise. At first, for the generation of the initial
stress state the whole model area is discretized using soil elements only. Subsequently, the
monopile is generated by replacing the soil elements located at the pile position by steel
elements and activating the contact conditions between pile and soil.
Then the vertical load is applied, and finally the horizontal load is applied and increased step
by step. The monopile elements were extended above the ground surface of the model in
order to realize different load combinations (horizontal forces and bending moments).

For a resultant horizontal load of 8 MN and a bending moment at sea-bed level of 240 MNm,
which is in the order of a possible design load for the considered large water depths, the
horizontal (bedding) stresses acting on the pile in the symmetry plane are shown in Fig. 3,
right. The characteristic loading behaviour of the pile with bedding stresses of opposite sign
above and below a point of rotation can be seen clearly. For the considered case the point of
rotation lies about 22 m below sea bed.

In Fig. 4, the numerically determined deflection lines for piles of two diameters (D = 2 m, D =
7.5 m) in homogeneous dense sand are plotted against the results computed according to the
API procedure. For the pile with 2 m diameter, results for a horizontal force of 3 MN agree
quite good in the case of high moment load (h/L = 1), whereas the deformations obtained by
the API method are clearly smaller with h/L = 0. For the pile with 7.5 m diameter, results
show in both cases relatively large deviations.
Of course, the finite element results do not necessarily reflect the reality, because assumptions
have to be made concerning initial stress state and material behaviour. Using calibrated
material parameters, good agreement may be obtained for both pile diameters in the case
h/L = 0. But, in that case the pile deformations for h/L = 1 will be significantly underestimated
by the API method.
The main reason for the deviations is probably the overestimation of the soil stiffness in large
depths by the API method. Considering the initial stiffness, the API method predicts for
medium dense sand stiffness values in the order of 500 MN/m2 in depths of around 20 m. This
is of particular influence on the results for high load levels, i. e. in the example given in Fig. 4
for h/L = 1 much more than for h/L = 0. Moreover, for a large-diameter pile the shearing resis-
tance in the pile tip area may play a more important role compared to a small-diameter piles.
It is thus concluded that in general the use of the API method for the design of large-diameter
monopiles is not suitable.
0 0
h/L = 0
h/L = 0

5 5 h/L = 1
h/L = 1

10 D=2m 10 D = 7.5 m
L = 18 m L = 20 m
Sand, dense Sand, dense
15 H = 3 MN 15 H = 8 MN
FEM FEM
API API
20 20
-0.5 0 5 10 15 20 -2 0 2 4 6 8
Displacement w in cm Displacement w in cm
Fig. 4: Exemplary comparison of the pile deflection according to API method and numerical
simulation for monopiles in dense sand.

Regarding the parametric studies, different pile diameters D, different pile lengths L and
different heights of point of load application h were numerically simulated. Diagrams
representing the pile head displacement w and the pile head rotation at sea-bed level as a
function of the horizontal load were determined. Thereby three soil profiles given in Fig. 5
were considered to idealize different soil profiles in the North Sea.

1 2 3

0 m Sea ground
-2 m Mud sand
Sand,
-8 m medium dense
Sand,
Sand,
medium
dense Boulder clay,
dense
semi-firm
-15 m

Sand, dense

Fig. 5: Idealised soil profiles considered in the numerical computations.

For each combination of pile geometry and soil profile, two diagrams were derived (H-w and
H- curves) as a result. These diagrams are represented in Figures 6, 7 and 8. If at a specific
site similar or comparable soil profiles exist, a pre-dimensioning of a monopile foundation for
static load design can be carried out on the basis of these diagrams.

The expected influence of the pile length and the height of the point of application of the
horizontal load is evident from the Figures given below. In homogeneous sand the
deformations for h/L = 1 are approximately double of the deformations for h/L = 0.2. For the
layered soil profile the ratio is even higher, which is mainly due to the relatively small
stiffness of the boulder clay layer.
10 10
h/L = 0.2 h/L = 0.2
Horizontal Force H in MN

Horizontal Force H in MN
8 8
h/L = 0.5
h/L = 0.5
6 6
h/L = 1
h/L = 1
4 4

2 2
dense dense
medium dense medium dense
0
0 5 10 15 20 0 0.2 0.4 0.6 0.8
Displacement w at sea ground level in cm Rotation O at sea ground level in

Fig. 6: Load-deformation curves for monopile D = 7.5 m, L = 20 m in medium dense and


dense sands (Soil profiles 1 and 2 with regard to Fig. 5).

16 16
h/L = 0.2 h/L = 0.2
Horizontal Force H in MN

Horizontal Force H in MN

12 12
h/L = 0.5 h/L = 0.5

8 h/L = 1 8 h/L = 1

4 4
dense dense
medium dense medium dense
0 0
0 5 10 15 20 0 0.2 0.4 0.6 0.8
Displacement w at sea ground level in cm Rotation O at sea ground level in

Fig. 7: Load-deformation curves for monopile D = 7.5 m, L = 30 m in medium dense and


dense sands (Soil profiles 1 and 2 with regard to Fig. 5).

16 16
Horizontal Force H in MN

Horizontal Force H in MN

h/L = 0.2 h/L = 0.2


12 12
h/L = 0.5 h/L = 0.5

8 h/L = 1 8
h/L = 1

4 4

0 0
0 5 10 15 20 0 0.2 0.4 0.6 0.8
Displacement w at sea ground level in cm Rotation O at sea ground level in

Fig. 8: Load-deformation curves for monopile D = 7.5 m, L = 30 m in layered soil (Soil


profile 3 with regard to Fig. 5).
Considering Figs. 6 and 7, the influence of the relative density of non-cohesive soils is
evident. Thus, a soil exploration with thorough determination of relative densities is to be
recommended for the design of monopile foundations.
Finally, comparing Fig. 6 and 7 the influence of the monopile lengths is elucidated. For a
monopile with 20 m length, the deformations with H = 10 MN are in the same order as with H
= 16 MN for a monopile with 30 m length. Thus, a 50% prolongation of the embedded pile
length leads in the considered case to an about 60% higher admissible load with regard to
serviceability of the foundation.

4 EFFECT OF CYCLIC LOADING

During the lifetime of an OWEP billions of loading cases induced by wave and wind actions
apply to the structure. Thus, the loading has to be classified as intensely cyclic and fatigue
design is of great importance.
Concerning the foundation structure, cyclic loading leads to an accumulation of permanent
displacements. According to the German offshore regulation (GL 1999), the effect of cyclic
loading of the foundation structure has to be taken into account. Unfortunately, concerning
monopiles no approved method exists to estimate the permanent displacements due to cyclic
loading. In the following, a short summary of the state of knowledge regarding this matter is
given.
Due to the API method, the factor A in equation (2) shall be set to 0.9 for cyclic loads. This
leads to moderate increases of the calculated displacements, which is elucidated by an
example calculation in Fig. 9.

16 16
Horizontal Force H in MN

Horizontal Force H in MN

h/L = 0 h/L = 0
12 12

8 h/L = 1 8 h/L = 1

4 4
static static
cyclic cyclic
0 0
0 2 4 6 8 0 0.1 0.2 0.3
Displacement w at sea ground level in cm Rotation O at sea ground level in

Fig. 9: Comparison of load-deformation curves for a monopile D = 7.5 m, L = 30 m in


medium dense sand for static and cyclic loading due to API (2000).

The API approach was derived by means of loading tests in which mostly less than 100 load
cycles were applied. In fact, strain accumulation does not stop reaching 100 cycles. Hettler
(1981) proposed the following equation for the displacement of a pile in sand loaded by N
cycles of the same horizontal load:
w N = w1 (1 + C N ln N ) (4)
Herein w1 is the displacement for static loading and CN is a factor which for sand lies in the
range of 0.20.
Long & Vanneste (1994) proposed a subgrade reaction method with linear increasing
subgrade modulus with depth, in which the moduli decrease with the number of load cycles
due to the following equation (see Fig. 10 left):
k s ( z ) = N t nh,1 z (5)
Herein t is a factor dependent on the pile installation method, the load characteristic (one- or
two-way loading) and on the relative density of the sand. For a driven pile with one-way
loading in medium dense sand t = 0.17 is recommended. With this value an example was
calculated. The results are given in Fig. 10. It was found that after 5000 cycles the
displacement at the pile top is more than 3 times the static displacement.

N = 1 50 500 5000 N = 1 50 500 5000


H 0 0
h

Depth z in m

10 10
z
L = 30 m

ks(z)
20 20

3 -0.17
h/L = 0 h/L = 1
ks,tip = 80 MN/m N H = 16 MN H = 16 MN
30 30
D = 7.5 m
-2 0 2 4 6 8 -5 0 5 10 15 20
Horizontal displacement w in cm

Fig. 10: Influence of cyclic one-way loading due to the method from Long & Vanneste (1994)
for a monopile D = 7.5 m, L = 30 m in medium dense sand.

In reality, the amplitude of the load is varying with the wave heights and the wind velocities.
For loads of varying amplitudes Lin & Liao (1999) proposed a strain superposition method to
determine the resultant permanent displacement. For that the following equations apply:
n

wGes = w1,1 1 + t ln N 1 + N k* (6.1)
k =2
1 w1, k

(1 + t ln N k )1
t w1,1
with N k* = e (6.2)

Herein w1,i is the static displacement due to a load Bi, which can be obtained by the
calculation of the static load-displacement curves presented in section 3.
However, due to Lin & Liao the displacements obtained with this method overestimate the
real displacements for large numbers of cycles. Additionally, also the directions of the loads
are varying. For two-way cyclic loading it is known that the cyclic displacements are
significantly smaller than for one-way loading due to densification effects in the soil (see
Hettler 1981, Long & Vanneste 1994). But at least the pile behaviour under general variation
of loading directions and amplitudes is an open question, which will be object of further
research.
5 CONCLUSIONS

The use of the API method for the computation of the deformations of large-diameter
monopile foundations for offshore wind energy plants cannot be generally recommended.
This applies to the design for static loads and particularly to the estimation of the influence of
cyclic loading.
For static loads, numerical investigations are recommended at present, as they were presented
in this paper. Of course, such investigations are complex and time-consuming. For
preliminary design steps diagrams can be helpful, which allow a simple determination of the
approximate pile deformations to be expected for a specific case.
The behaviour of monopile foundations under cyclic loading of varying amplitude and
direction is up to now not well understood and must be a subject of future research.

The results presented in this paper were obtained in the framework of the FORWIND research
group funded by the Government of the federal state of Lower Saxony, Germany. The support
is gratefully acknowledged.

REFERENCES

[1] Abaqus 2004. Users Manual. Version 6.4.


[2] Achmus, M. & Abdel-Rahman, K. 2004. Numerische Untersuchung zum Tragverhalten
horizontal belasteter Monopile-Grndungen fr Offshore-Windenergieanlagen. 19. Christian
Veder Kolloquium, Graz/sterreich.
[3] API 2000. American Petroleum Institute. Recommended Practice for Planning, Designing
and Constructing Fixed Offshore Platforms- Working Stress Design, API Recommended
Practice 2A-WSD (RP2A-WSD), 21st edition, Dallas.
[4] DNV 2004. Det Norske Veritas. Design of Offshore Wind Turbine Structures. Offshore
Standard, Norway.
[5] EAU 2004. Empfehlungen des Arbeitsausschusses Ufereinfassungen Hfen und
Wasserstraen. Verlag Ernst & Sohn, Berlin.
[6] GL 1999. Germanischer LIoyd. Rules and Regulations, Offshore Wind Energy
Converters. Hamburg, Germany.
[7] Hettler, A. 1981. Verschiebungen starrer und elastischer Grndungskrper in Sand bei
monotoner und zyklischer Belastung. Verffentlichungen des Instituts fr Bodenmechanik
und felsmechanik der Universitt Karlsruhe, Heft 90.
[8] Lesny, K.; Richwien, W.; Wiemann, J. 2002. Grndungstechnische Randbedingungen fr
den Bau von Offshore-Windenergieanlagen in der Deutschen Bucht. Bauingenieur Band 77,
S. 431 438.
[9] Lin, S.-S.; Liao, J.-C. 1999. Permanent strains of piles in sand due to cyclic lateral loads.
ASCE Journal of Geotechnical and Geoenvironmental Engineering, Sept.
[10] Long, J.H.; Vanneste, G. 1994. Effects of cyclic lateral loads on piles in sand. ASCE
Journal of Geotechnical Engineering, Vol. 120, No. 1.
[11] Lpile 2000. Users manual, Version Lpile plus 4.0.
[12] Reese, L. C.; Cox, W. R.; Koop, F. D. 1974. Analysis of Laterally Loaded Piles in Sand.
Proceedings of the VI Annual Offshore Technology Conference. Dallas.
[13] Wiemann, J.; Lesny, K. 2004. Evaluation of the Pile Diameter Effects on Soil-Pile
Stiffness. Proceedings of the 7th German Wind Energy Conference (DEWEK),
Wilhelmshaven.

You might also like